Read time: 9 minutes

What you’ll learn:

• How caffeine keeps disrupting sleep for hours longer than most people assume
• Why alcohol can help you fall asleep and still wreck the second half of your night
• What the research shows about eating close to bedtime
• Whether tryptophan-rich foods and “sleepy” snacks are worth the hype
• Where diet fits into insomnia treatment, and where it doesn’t

How What You Eat and Drink Affects Your Sleep

A woman in Kelowna sits down at her kitchen table at 9 p.m. with a glass of wine, telling herself it’s the one thing that reliably gets her to sleep. Across town, a man finishes his last coffee at 3 p.m. because he read that’s the cutoff, then wonders why he’s still staring at the ceiling at midnight. Both are responding to real physiology. Neither has the full picture.

Diet and sleep are connected in ways that are measurable, but the size of the effect varies enormously depending on what you’re eating or drinking, and when. Some of it is dramatic. Some of it barely moves the needle. This post works through the evidence food by food, so you can tell the difference.

Caffeine keeps working long after the cup is empty

Caffeine’s reputation for disrupting sleep is well earned, and the research shows the disruption lasts far longer than the “avoid it after lunch” advice implies.

A 2013 study by Drake and colleagues gave participants 400 mg of caffeine (roughly the amount in three to four cups of brewed coffee) at 0, 3, or 6 hours before their normal bedtime. Even at the 6-hour mark, caffeine reduced total sleep time by 41 minutes compared to placebo, and every dosing time disrupted objective sleep measures on polysomnography. A 2024 crossover trial by Gardiner and colleagues found that 400 mg consumed within 4 hours of bedtime reduced perceived sleep quality by roughly a third, while a 100 mg dose (about one cup) caused no measurable disruption when taken 4 or more hours out.

A 2023 systematic review pooling 24 studies puts numbers on the average damage: caffeine cuts total sleep time by about 45 minutes, drops sleep efficiency by 7%, adds 9 minutes to sleep onset latency, and adds 12 minutes of wake time after sleep onset. The same review calculated that to avoid measurable sleep loss, a standard cup of coffee needs roughly 8.8 hours of lead time before bed. A pre-workout supplement, with its higher caffeine dose, needs over 13 hours.

The practical takeaway isn’t “cut caffeine entirely.” It’s that the common cutoffs (noon, 2 p.m.) are built for average metabolism and average sensitivity, and insomnia patients tend to be more caffeine-sensitive than average, not less. If sleep is already fragile, an earlier and firmer cutoff does more good than the general population advice suggests.

Alcohol trades a faster start for a rougher night

Alcohol is the substance people most often reach for specifically to sleep better, and it’s also the one where the research most clearly contradicts that intention.

According to a 2025 systematic review and meta-analysis, at low doses, roughly two standard drinks, alcohol delays and shortens REM sleep without meaningfully changing how long it takes to fall asleep. At higher doses, five or more standard drinks, sleep onset does shorten, but the second half of the night pays for it: more slow-wave sleep early on, then a collapse into fragmented, lighter sleep later, with reduced REM that doesn’t rebound the way some older research suggested it would. A 2024 study using three consecutive nights of pre-sleep alcohol found the same pattern held night after night, with REM sleep consistently reduced and no adaptation over time.

This explains a pattern many insomnia patients recognize: the wine or nightcap seems to work, because sleep onset genuinely does come faster. What it doesn’t show up in is how the person feels at 3 a.m., or how unrefreshed they are in the morning. Alcohol is solving the wrong problem. It shortens the time to fall asleep while degrading the sleep that follows, which is the opposite of what someone with fragmented, non-restorative sleep needs.

Late dinners push sleep onset later

Meal timing gets less attention than caffeine or alcohol, but the data on eating close to bedtime is more consistent than most people expect.

A 2024 analysis by Yan and colleagues of over 7,000 NHANES participants found that later meal timing, whether measured by first meal, midpoint, or last meal of the day, was associated with worse scores on the Pittsburgh Sleep Quality Index. A tightly controlled 2026 crossover trial by Enomoto and colleagues went further: healthy young women who ate dinner one hour before bed had shorter total sleep time, lower sleep efficiency, more wake time after sleep onset, and more nighttime arousals than when they ate the same dinner five hours before bed. A separate study of night-shift workers found that every hour the last meal moved closer to sleep onset was associated with a meaningful increase in subsequent sleep duration when workers slept during the day, though higher fat and carbohydrate content in that meal independently lengthened the time it took to fall asleep.

The mechanism isn’t mysterious. Eating is a strong circadian cue, and eating during the body’s biological night works against the same internal clock that’s trying to wind down for sleep. Fat and carbohydrate content matter too, likely through digestion, reflux risk, and shifts in insulin, leptin, and ghrelin, all of which run on their own daily rhythms.

None of this means dinner has to happen at 5 p.m. It means a meal eaten within an hour of lights-out is doing measurably more to disrupt sleep than the same meal eaten three or four hours earlier, and for anyone already fighting fragmented sleep, that gap is worth protecting.

Tryptophan-rich foods have a small, inconsistent effect

Turkey, warm milk, and other tryptophan-rich foods carry a persistent reputation as natural sleep aids, and the underlying biology is real: tryptophan is a precursor to serotonin, which is in turn a precursor to melatonin. The evidence that eating tryptophan-rich food changes sleep outcomes is thinner and more mixed than the reputation suggests.

A study by Fukushige and colleagues on Japanese university students found that a tryptophan-rich breakfast raised evening melatonin only when combined with bright light exposure during the day. Tryptophan alone, without the daytime light, didn’t move melatonin by a meaningful margin. A separate, Nestlé-funded trial testing an evening supplement combining tryptophan with a glucose-lowering mulberry leaf extract did produce a measurable, if modest, improvement: about four fewer minutes to fall asleep and a small bump in next-morning mood. Systematic reviews on milk and dairy specifically report improved scores on standard sleep questionnaires, but the individual trials are small, heterogeneous, and the authors of more than one review flagged the need for larger studies before drawing firm conclusions.

The honest summary is that tryptophan-rich foods are not harmful, may help a little at the margins, and are nowhere close to a treatment for insomnia on their own. Anyone told that a glass of milk or a turkey sandwich will resolve months of fragmented sleep has been sold something the evidence doesn’t support.

A few patterns worth holding onto

Pulling the research together, a few consistent patterns hold up:

• Caffeine cutoff, moved earlier than the general advice. Given caffeine’s individually variable half-life, an early-afternoon cutoff is a reasonable default for anyone with insomnia, with room to experiment.
• Alcohol treated as a sleep disruptor, not a sleep aid, regardless of how convincing the faster sleep onset feels in the moment.
• A buffer of at least two to three hours between the last meal and bedtime, longer if the meal is high in fat or refined carbohydrate.
• A diet built around whole grains, vegetables, and adequate fibre, which the broader nutrition literature associates with better sleep quality, likely through blood sugar stability and reduced inflammation rather than any single “sleep nutrient.”

None of these are dramatic interventions, and none will resolve chronic insomnia by themselves. They’re supporting conditions, the same category as room temperature or a consistent wake time.

Where diet fits into insomnia treatment

This is the distinction that matters most, and it’s the one that gets blurred in most articles on this topic. Sleep hygiene, which includes diet and drink timing, is a supporting layer. It is not a treatment for insomnia. Cognitive behavioural therapy for insomnia works on the mechanisms that sustain chronic insomnia: the conditioned arousal around bed, the mismatch between time in bed and actual sleep need, and the anxious monitoring that keeps the nervous system on alert at night. Diet changes can support that work. They can’t substitute for it.

There’s also a trap worth naming directly. Treating food and drink timing as a strict rulebook, tracking caffeine to the minute, refusing dinner invitations that run late, panicking over a single glass of wine, can itself become a source of the hypervigilance that keeps insomnia going. The goal is a few sensible defaults, held loosely, not another set of nighttime rules to monitor and enforce.

If diet and drink adjustments haven’t moved the needle after a few weeks, that’s not a sign to tighten the rules further. It’s a sign the problem sits somewhere else, and that’s worth an assessment rather than more restriction.

TL;DR

• Caffeine disrupts sleep for far longer than commonly advised. Even 400 mg taken 6 hours before bed cut sleep time by 41 minutes in controlled research; an early-afternoon cutoff is a safer default for anyone with insomnia.
• Alcohol shortens time to fall asleep at higher doses but fragments and shortens sleep later in the night, with reduced REM that doesn’t rebound. It functions as a sleep disruptor, not a sleep aid.
• Eating close to bedtime measurably reduces total sleep time and sleep efficiency in controlled trials; a two- to three-hour buffer before bed, longer for heavy or high-fat meals, is well supported.
• Tryptophan-rich foods show a small and inconsistent effect on melatonin and sleep onset, helpful at the margins but not a treatment.
• Diet is a supporting layer, not a treatment. For chronic insomnia, CBT-I addresses the mechanisms that keep sleep fragmented; diet changes support that work without replacing it.

Struggling with insomnia and not sure whether it’s your diet, your schedule, or something else entirely? Book an assessment with BC CBT-I to find out what’s keeping you up.

About the author

Graeme Thompson, MA, RCC, is a Registered Clinical Counsellor (BCACC #21951) and the founder of BC CBT-I, a virtual practice providing evidence-based insomnia treatment across British Columbia using CBT-I and ACT-I.

References

Drake, C., Roehrs, T., Shambroom, J., & Roth, T. (2013). Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. Journal of Clinical Sleep Medicine, 9(11), 1195–1200.

Enomoto, M., Kitamura, S., Kunieda, T., et al. (2026). Effects of later dinner timing on subsequent metabolic function and nocturnal sleep in healthy young women. Journal of Physiological Anthropology, 45, 12.

Fukushige, H., Fukuda, Y., Tanaka, M., Inami, K., Wada, K., Tsumura, Y., Kondo, M., Harada, T., Wakamura, T., & Morita, T. (2014). Effects of tryptophan-rich breakfast and light exposure during the daytime on melatonin secretion at night. Journal of Physiological Anthropology, 33, 33.

Gardiner, C. L., Weakley, J., Burke, L. M., Fernandez, F., Johnston, R. D., Leota, J., Russell, S., Munteanu, G., Townshend, A., & Halson, S. L. (2024). Dose and timing effects of caffeine on subsequent sleep: A randomized clinical crossover trial. Sleep, 48(4), zsae230.

Gardiner, C., Weakley, J., Burke, L. M., Roach, G. D., Sargent, C., Maniar, N., Townshend, A., & Halson, S. L. (2023). The effect of caffeine on subsequent sleep: A systematic review and meta-analysis. Sleep Medicine Reviews, 69, 101764.

Gardiner, C., Weakley, J., Burke, L. M., Roach, G. D., Sargent, C., Maniar, N., Huynh, M., Miller, D. J., Townshend, A., & Halson, S. L. (2025). The effect of alcohol on subsequent sleep in healthy adults: A systematic review and meta-analysis. Sleep Medicine Reviews, 80, 102030.

McCullar, K. S., Barker, D. H., McGeary, J. E., Saletin, J. M., Gredvig-Ardito, C., Swift, R. M., & Carskadon, M. A. (2024). Altered sleep architecture following consecutive nights of presleep alcohol. Sleep, 47(4), zsae003.

Nogueira, L. F. R., Pellegrino, P., Cipolla-Neto, J., Moreno, C. R. C., & Marqueze, E. C. (2021). Timing and composition of last meal before bedtime affect sleep parameters of night workers. Clocks & Sleep, 3(4), 605–616.

Soon, C. S., Thota, R., Owen, L., Tian, L., Martin, F.-P., Mantantzis, K., Cherta-Murillo, A., Campos, V. C., Chkroun, C., Lavalle, L., Hartweg, M., St-Onge, M.-P., Chee, M. W. L., & Darimont, C. (2025). Mulberry leaf extract combined with tryptophan improves sleep and post wake mood in adults with sleep complaints: A randomized cross-over study. European Journal of Nutrition, 64(3), 124.

Yan, L.-M., Li, H.-J., Fan, Q., Xue, Y.-D., & Wang, T. (2024). Chronobiological perspectives: Association between meal timing and sleep quality. PLOS ONE, 19(8), e0308172.